=Paper=
{{Paper
|id=Vol-2790/paper16
|storemode=property
|title=
Big Data Environmental Monitoring System in Recreational Areas
|pdfUrl=https://ceur-ws.org/Vol-2790/paper16.pdf
|volume=Vol-2790
|authors=Alexander Volkov,Andrey Kopyrin,Natalya Kondratyeva,Sagit Valeev
|dblpUrl=https://dblp.org/rec/conf/rcdl/VolkovKKV20
}}
==
Big Data Environmental Monitoring System in Recreational Areas
==
https://ceur-ws.org/Vol-2790/paper16.pdf
Big Data Environmental Monitoring System in
Recreational Areas
Alexander Volkov1, Andrey Kopyrin1, Natalya Kondratyeva1 and Sagit Valeev1
1 Sochi State University, Sochi, 354008, Russia
vss2000@mail.ru
Abstract. The paper discusses the architecture of the environmental monitoring
system of the recreational area. It is assumed that the system uses big data tech-
nologies to process heterogeneous information. Information in various formats is
obtained using remote sensing of the earth, aerial photography, shooting from
unmanned aerial vehicles, and internal monitoring of the state of objects of the
recreational zone. The area of pollution and the volume of emissions and, conse-
quently, the concentrations created by them are determined by experts consider-
ing the available data. The necessity of considering the dynamics of the ecologi-
cal state based on the available data generated by weather services and remote
sensing data of the earth in various spectral ranges is shown. Also, within the
framework of the monitoring system, statistical data and results of modeling the
ecological state of the recreational zone are used. Based on the analytics of big
data, it is supposed to build a forecast of thermal emissions and the amount of
harmful substances in the atmosphere carried by air masses.
Keywords: Environmental Monitoring, Big Data, Multi-level Information Col-
lection System, Databases, Data Aggregation.
1 Introduction
As it is known, the main purpose of the recreational zone is to restore the health of the
population and ensure the preservation of the labor capital of any state. Assessment of
the pollution of the recreational area, considering the scale of environmental pollution,
allows you to control the ecological situation in recreational areas within the region.
Assessment of the state of complex systems is based on a hierarchical collection of
models of various details [1]. When building a monitoring system for the ecological
state of a recreational zone, models of various details are used, based on the analysis of
available information.
Thermal emissions and emissions of harmful substances carried by air masses sig-
nificantly affect the ecological state of coastal recreational zones. These circumstances
affect the level of recreational services provided, as well as the ecological state of the
resort area. Predicting the ecological state and reducing harmful emissions of various
natures is an extremely urgent scientific and practical task [2, 3].
Determining the area of these zones of increased concentration of harmful sub-
stances allows us to qualitatively assess the level of pollution of the zone based on
Copyright © 2020 for this paper by its authors. Use permitted under Creative
Commons License Attribution 4.0 International (CC BY 4.0).
171
expert estimates and big data analytics. The area of zones with an elevated level of
harmful substances at different times is significantly different. This is due to several
reasons, among which the climatic conditions of the considered recreational region and
the trajectory of the air masses transporting harmful substances over long distances [4-
7].
From this we can conclude that it is necessary to take into account the dynamics of
the ecological state, taking into account the available data generated by weather ser-
vices and data from remote sensing of the earth in various spectral ranges.
To assess the permissible concentration of harmful emissions, stationary systems for
measuring air quality and monitoring based on aerial photography and satellite images
are used.
The area of pollution and the volumetric consumption of emissions and, conse-
quently, the concentrations created by them are determined by experts considering the
available data. In total, thermal emissions into the environment and emissions of harm-
ful substances from megacities lead to a significant deterioration of the ecological state
of the resort area, which is quite far from the places of emission.
To analyze and make an effective forecast, the problem of integrating systems for
collecting and processing a large amount of data is considered.
Fig. 1 shows a map of emissions of harmful substances by megacities in the resort
areas of Spain in the winter, shows areas of increased concentration of harmful sub-
stances on the coast Z1 and Z2. Solid arrows show the vectors of movement of air masses
in this period.
The determination of the area of these zones allows a qualitative assessment of the
level of pollution based on expert estimates.
Fig. 1. Map of the distribution of harmful emissions in recreational areas of the coast of Spain in
winter.
As follows from Fig. 2, the area of zones with an increased level of harmful substances
in the high season is much larger than in the low season. This is due to several reasons,
among which the climatic conditions of the considered recreational region and the in-
crease in demand for services in certain periods of the year. From this we can conclude
that it is necessary to take into account the dynamics of the ecological state and possible
172
emissions of harmful substances associated with heating and cooling air in the prem-
ises, taking into account the zonal climatic conditions of the recreation area.
The effectiveness of solving the problem of improving energy efficiency, and
thereby reducing harmful emissions during heating and cooling can be achieved using
modern technologies and methods. Among them are the reconstruction of the existing
construction fund, the district heating network and the introduction of new generation
cooling systems [8-13].
Fig. 2. Map of the distribution of harmful emissions in recreational areas of the coast of Spain in
the summer.
It should be noted that a feature of the recreational zones of the seacoasts is the presence
in the general fund of recreational infrastructure facilities of many small private hotels
with limited opportunities for the application of innovative energy-saving technologies.
Another feature of the recreational areas is the specificity of the transport infrastructure,
which is overloaded during peak seasons, which leads to traffic jams and increased
emissions of harmful substances.
The concentration of emissions significantly depends on the wind rose, as well as on
the time of day and time of year. Thus, it should be noted that in the summer season
additional active measures are required to control harmful emissions by vehicles.
Thus, the task of creating an information system for monitoring the level of pollution
of a recreational zone based on the use of multilevel data analysis is urgent. A feature
of the proposed solution is the collection and storage of data from all available infor-
mation sources and their integration for a step-by-step analysis of pollution sources, as
well as taking into account the dynamics of changes in the level of pollution, taking
into account the movement of air masses transporting harmful substances from remote
industrial regions [14-17].
The principles of designing an information system for environmental monitoring of
the recreational zone, which uses data from remote sensing of the earth, aerial photog-
raphy, surveys from unmanned aerial vehicles, internal monitoring of the state of the
objects of the recreational zone, as well as statistical data, were discussed in [18].
173
Next, we will consider the development of this idea based on a three-level hierar-
chical information system for collecting heterogeneous information about the ecologi-
cal state of the recreational zone and algorithms for filtering noise using neural network
technologies.
2 The Task of Analyzing the Amount of Harmful Emissions
Solving the problem of analyzing the dynamics of changes in air pollution in recrea-
tional areas, taking into account seasonal temperature fluctuations and the amount of
emissions of harmful substances from vehicles, taking into account the influence of the
direction of movement of air masses, depending on the season, carrying harmful sub-
stances from large cities and industrial enterprises, allows to determine the harm ap-
plied to the recreational area.
It is assumed that data are collected both at the regional level and at the level of
individual buildings, which will make it possible to determine the level of pollution and
their main causes at various levels of the hierarchy, which, in total, can provide an
opportunity to control the situation [19-20].
As you know, processing the results of aerial photography in the infrared spectrum
allows you to localize sources of heat loss, including in heating networks. The use of
small-sized aircraft makes it possible to collect statistical data taking into account the
time of day with higher accuracy and at lower material costs.
The problem of determining the area of pollution can be solved by determining the
sum of areas of pollution sources. In this case, numerical integration is not assumed.
The search can be reduced to the analysis of images with the determination of the total
area of contamination by the images themselves.
The total level of pollution taking into account the emissions of heat and harmful
substances can be estimated using the following expressions:
S H xdy S H1 ... S Hm , S H S HS
m H
S C xdy S C1 ... S Cn , S C S CS
n C (1)
S O xdy S O1 ... S Ok , S O S OS
k O
R k S H l SC m SO
where R is the overall level of pollution; SH is emissions of harmful substances associ-
ated with heating; SC is emissions of harmful substances associated with cooling; SO is
emissions associated with traffic flows; k, l, m are weighting factors, H– heating, C –
cooling, O – other loses.
Threshold values of permissible levels of contamination for (1) are determined on
the basis of regulatory documents of various departments and organizations.
The choice of the values of the weight coefficients k, l, m depends on many factors
determined by expert assessments, various standards, etc.
174
The volume of emissions is characterized by the area and concentration of harmful
substances. It is assumed that the data collection performed by stationary measurement
systems allows us to obtain an average value taking into account the season.
Fig. 3 presents a map of emissions of harmful substances in high season. To assess
the permissible concentration of harmful emissions, stationary air quality measuring
systems are used.
Fig. 3. Emission distribution map.
Fig. 4. High season emissions.
Area of pollution Z as well as volumetric emissions consumption and, therefore, the
concentrations created by them are determined by experts considering the available data
(Fig. 4). In aggregate, thermal emissions into the environment and emissions of harmful
substances by transport systems lead to a significant deterioration of the ecological state
of the resort area. To analyze and make an effective forecast, it is necessary to actively
use systems for collecting and processing a large amount of data.
To solve this problem, it is proposed to use a multi-level system for collecting infor-
mation on the environmental characteristics of the recreational area [4, 5]. The data
collection and processing system integrates data arrays obtained by processing infor-
175
mation from satellite images, distributed monitoring data on the thermal state of infra-
structure facilities, as well as the concentration of harmful substances in the resort area
at different times of the day and time of the year.
Considering the wind direction makes it possible to improve the quality of the fore-
cast, since it considers the level of pollution, the change in which is caused by air
masses carrying harmful substances.
3 The Architecture of the Environmental Monitoring System of
the Recreational Area
The data aggregation system includes data processing subsystems obtained on the basis
of image analysis and weather observation data. Thus, the problem under consideration
is primarily associated with the organization of the collection of information on the
ecological state of resort facilities. The complexity of solving the problem under con-
sideration is due to the need to collect process and store heterogeneous information, as
well as the need to solve the problem associated with the development of a forecast of
the ecological state of the region.
To solve the problems of processing heterogeneous information of large volume,
certain hopes are laid on their solution using big data technologies. This is due to the
successful practical development of new data transfer technologies, the implementation
of data management systems based on NoSQL technologies, as well as the successful
implementation of distributed data processing based on cloud computing.
In the framework of the use of big data technologies, the application of algorithms
for determining energy losses and pollution levels based on the results of image analysis
and statistical results of environmental modeling is promising. In this regard, the devel-
opment of a system for collecting data on energy consumption can ensure the efficiency
of the procedure for developing a plan for the generation of thermal energy for heating
and cooling, and optimal plans for loading transport arteries in the region, taking into
account the wind rose and climatic conditions.
When collecting data by aircraft flying at different heights, the level of detail of the
images obtained changes, which leads to the need to synchronize the received data in
time and coordinates. At the same time, the formats for presenting graphical infor-
mation and data from various sources can differ significantly.
Fig. 5 presents the generalized architecture of a three-level data collection system
for monitoring the energy consumption used in heating, cooling facilities of a recreation
area in various climatic conditions and collecting data on air quality.
A feature of the organization of computational processes and data storage in the sys-
tem under consideration is the need to process data on the thermal state of the spa zone,
the thermal state of spa resort facilities and the internal thermal state of objects pre-
sented in various formats.
The complexity of solving this problem is due to the need for reliable localization of
objects and linking all available information about losses to these objects.
In Fig. 5 at the city level L1 (where Oi, i = 1…m, are areas of different energy losses),
based on the analysis of data on excess electricity consumption, heat losses in heating
176
networks, the concentration of harmful substances at the current time, determined by
the state of the city's transport routes, it is possible to determine zones that are unfavor-
able from the point of view of the ecological condition.
Fig. 5. Generalized scheme of a three-level system for collecting data from monitoring results
Data obtained on the basis of analysis of power consumption, temperature measure-
ments, wind directions, measurements of the concentration of harmful substances with
reference to a city map are stored in the DB_L1 database.
To obtain information on the state of the characteristics of the ecological state of the
region, statistical information is collected from urban environmental services, as well
as information obtained on the basis of aerial photography and images in various spec-
tral ranges obtained using meteorological satellites. The analysis results at this level
make it possible to determine the level of pollution in various areas of the recreational
zone (Rj, j=1…k). Data at this level is aggregated in the DB_L2 database.
Sources of pollution, which can be attributed to industrial enterprises, transport sys-
tems of megalopolises, to a large extent currently affect the ecological state of recrea-
tional areas. This is caused by the transport of harmful substances by air masses over
long distances (in Fig. 5 PC is a pollution cloud, Wd is a wind direction). Information
on the direction of air masses is presented at the L3 level, data on the speed and direc-
tion of air masses is in the database BD_L3. The data sources are L1 and L2 weather
stations. The results of satellite imagery analysis can also be used. Data considering the
heterogeneity of formats, the time of their receipt are placed in the DB_I database.
These data are used to build the PS environmental forecasting system. When a request
177
for forecasting Q arrives, the forecasting system, using the integrated data on the current
state of the system L1, L2, L3, issues forecast P.
Several states of the ecological system should be distinguished: normal, harmful to
health, and critical. Depending on the results of the analysis, the consumers of this in-
formation may be residents of the recreational zone or the relevant state and municipal
services.
As noted earlier, data integration means the use of all available data on the ecological
state of objects in the recreational zone and the zone itself, as well as solving the prob-
lem of linking them to the objects under consideration and to the zone itself. Images
obtained using various data collection tools have different resolutions and details. They
were obtained at different times and can display different spectral components of the
object state. Obtaining a general picture requires solving problems of semantic pro-
cessing and using various algorithms for image analysis.
Based on the data obtained, it is proposed to construct a textbook for training neural
networks, with the help of which it is supposed to obtain a forecast.
Determining the area of objects under study using software systems for analyzing
space images involves a large proportion of manual processing using graphic tools built
into these systems. When determining the level of pollution on images with a high res-
olution, it is necessary to solve the problem of assessing the total area of pollution
zones. It should be noted that the quality of images depends on many factors: illumina-
tion, cloudiness, etc.
The data aggregation system includes subsystems for image processing, determining
the level of thermal state of objects, and localization of objects. Aggregated information
is placed in the database DB_I.
In the case of processing a video series, the problem of image recognition based on
frame extraction arises.
Fig. 6 shows an image analysis algorithm that allows you to determine the quality of
satellite images depending on the cloud cover.
The proposed algorithm consists of frame extracting, pre-processing stage, training
stage and testing stage. In Fig. 6, TSN segment is a segment of the temporal segment
network (TSN), a novel framework for video-based action recognition. It refers to the
idea of long-range temporal structure modeling.
It includes a sparse temporal sampling and video-level supervision. Data augmenta-
tion is a common technique to improve results and avoid overfitting during model train-
ing. To solve the problem of image analysis, a convolutional neural network is used
(CNN) [18, 19].
Thus, the problem under consideration is primarily associated with the organization
of the collection of information on the thermal state of resort facilities and the state of
air quality.
178
Fig. 6. Algorithm of image classification (Cloudy/No Cloudy)
The complexity of the solution of the problem under consideration is due to the need
to store heterogeneous information and the need to solve the problem of localizing re-
sort facilities, determining heat losses in different climatic conditions, taking into ac-
count the peculiarities of energy use for various needs.
4 Conclusion
The problem of organizing the collection of information on the ecological state of the
resort area is considered. The complexity of the solution is due to the need to organize
179
the storage of heterogeneous information and solve the problem of localization of resort
facilities, determine heat losses in various climatic conditions, taking into account the
peculiarities of the use of energy resources, as well as the problem of analyzing the
state of air quality in the dynamics of the quality of the environment. A multilevel data
collection system using data from remote sensing of the Earth, aerial photography and
surveying from unmanned aerial vehicles, as well as data from internal monitoring of
both the thermal state of objects in the recreation area and air quality are discussed.
It is assumed that the proposed solutions may be in demand when assessing the harm
caused to the recreational area by industrial emissions and insufficient attention to en-
ergy-saving technologies for heating and cooling. The results of the analysis can serve
as a factual base when searching for solutions at various levels of decision-making,
including the regional level.
References
1. Thalheim, B.: Models for Communication, Understanding, Search, and Analysis. In: Pro-
ceedings of the XXI International Conference on Data Analytics and Management in Data
Intensive Domains (DAMDID/RCDL 2019), pp. 3-18 (2019)
2. Overall heat transfer loss from buildings - transmission, ventilation and infiltration,
https://www.engineeringtoolbox.com/heat-loss-buildings-d_113.html
3. Theodore, L., Behan, K.: Introduction to Optimization for Chemical and Environmental En-
gineers, 1st ed. CRC Press (2018)
4. Top six places for energy losses in commercial buildings. https://www.ec-
donline.com.au/content/test-measurement/article/top-six-places-for-energy-losses-in-com-
mercial-buildings-259097222
5. Building Envelope: How to Avoid Energy Loss. https://www.facilitiesnet.com/energyeffi-
ciency/article/Building-Envelope-How-to-Avoid-Energy-Loss--9428
6. Aerial Infrared Inspection. http://nationwidedrones.co.uk/drone-infrared-inspection
7. Making Your Facilities A Safer Place. http://preciseir.com/
8. Leskovec, J., Rajaraman, A., Ullman, J.D.: Mining of massive datasets. Cambridge Univer-
sity Press (2014)
9. Bohlouli, M., Schulz, F., Angelis, L., Pahor, D., Brandic, I., Atlan, D., Tate, R.: Towards an
integrated platform for Big Data analysis. In: Integration of Practice-Oriented Knowledge
Technology: Trends and Prospectives. Springer, Berlin, pp. 47–56 (2013)
10. Fagin, R., Kolaitis, P.G., Miller, R.J., Popa, L.: DATA exchange: semantics and query an-
swering, in Proc. of the 9th Int. Conf. on Database Theory (ICDT 2003), pp. 207–224 (2003)
11. Jacobs, B.: Categorical Logic and Type Theory. In: Studies in Logic and the Foundation of
Mathematics. 141. Elsevier, Amsterdam (1999)
12. Jones, M., Schildhauer, M., Reichman, O., Bowers, S.: The new bioinformatics: integrating
ecological data from the gene to the biosphere. In: Annu. Rev. Ecol. Evol. Syst. 37(1), pp.
519–544 (2006)
13. Labrinidis, A., Jagadish, H.V.: Challenges and opportunities with Big Data. In: Proc. VLDB
5(12), pp. 2032–2033 (2012)
14. Lenzerini, M.: Data integration: a theoretical perspective. In: Proc. of the 21st ACM SIGACT
SIGMOD SIGART Symp. on Principles of Database Systems (PODS 2002), pp. 233–246
(2002)
180
15. Lenzerini, M., Majki´c, Z.: First release of the system prototype for query management.
Semantic webs and agents in integrated economies, D3.3, IST-2001-34825 (2003)
16. Lenzerini, M., Majki´c, Z.: General framework for query reformulation. Semantic webs and
agents in integrated economies, D3.1, IST-2001-34825, February (2003)
17. Sazontev, V.: Methods for Big Data Integration in Distributed Computation Environments.
In: Proceedings of XX International Conference “Data Analytics and Management in Data
Intensive Domains” (DAMDID/RCDL’2018), pp. 239-244 (2018)
18. Volkov, A.N., Kopyrin, A.S., Kondratyeva, N.V., Valeev, S.S.: Ecological Monitoring In-
formation System in Recreational Zones, Scientific Journal Engineering System and Con-
structions, 1(38), pp.20-24 (2020) (in Russian)
19. Kondratyeva, N.V., Valeev, S.S.: Simulation of the life cycle of a complex technical object
within the concept of Big Data. In: CEUR Proceedings of 3rd Russian Conference Mathe-
matical Modeling and Information Technologies, pp. 216-223 (2016)
20. Volkov, A., Kopyrin, A., Kondratyeva, N., Valeev, S.: Multilevel Data Acquisition System
of Energy Losses in Recreation Areas. In: 2019 Twelfth International Conference "Manage-
ment of large-scale system development" (MLSD’2019), Moscow, Russia, pp. 1-4 (2019)
181